Credit & treasury · finance workload

compute.covenant.v1

Compute LTM Adjusted EBITDA, net leverage, and fixed charge coverage from four quarters of financials under a credit agreement's own definitions — capped restructuring add-backs (greater/lesser of a fixed amount and a percentage), litigation add-backs allowed or excluded, cash netting capped or prohibited — and state compliance with each covenant (15 borrower scenarios).

No published results for this workload yet — results appear here with the next edition.

Rate-card unit

Test a quarter's covenant package — rates on the rate card are the cost per successful outcome scaled to this unit.

Pass criteria

Adjusted EBITDA within ±$2k, both ratios within ±0.02, and both compliance determinations exactly correct. Grading is deterministic; a model output either passes or fails, and infrastructure errors are excluded and rerun rather than counted against the model.

Prompt (verbatim)

Every model receives exactly this prompt with the item's fields substituted for the {placeholders} — no model-specific tuning.

You are testing quarterly covenant compliance for a borrower. Apply the
definitions in the credit agreement excerpt below exactly as written — they
control over any standard convention. All amounts are in $ thousands; the
measurement period is the four quarters provided (LTM); balance-sheet tests
use the Q4-end balances.

Compute:
  adjusted_ebitda_ltm — Consolidated Adjusted EBITDA for the LTM period,
    rounded to the nearest thousand (an integer)
  leverage_ratio — Consolidated Total Net Debt / Consolidated Adjusted
    EBITDA, rounded to two decimals
  fixed_charge_coverage — the Fixed Charge Coverage Ratio, rounded to two
    decimals
  leverage_compliant / fcc_compliant — whether each covenant in Section 6.12
    is satisfied

Respond with ONLY a JSON object, e.g.
{{"adjusted_ebitda_ltm": 61250, "leverage_ratio": 3.42,
  "fixed_charge_coverage": 1.31, "leverage_compliant": true,
  "fcc_compliant": false}}

{agreement}

Financial data (CSV, $ thousands):
{financials}

Dataset

Datasets are generated synthetically from seeded generators; the seed is frozen per edition and held out until the run window closes. Generator source hash (SHA-256): 077287bc5d568713ec056adc40d7900b8620f824a3af2b4b94b268e986fca2b8. See the methodology §7 for the verifiability protocol and any edition page for the raw graded logs.